View Single Post
Old 2020-05-05, 19:17   #64
If I May
chalsall's Avatar
"Chris Halsall"
Sep 2002

928610 Posts

Originally Posted by SethTro View Post
It doesn't feel good to spend time and effort working on something then have someone call it bad. The use of very feels unrestrained.
Juggling some things at the moment, but...

Please never take offense from feedback here. It's never personal -- we're just trying to come to convergence.

At the end of the day, having a "Proof of Work" code-path is going to be a hard sell if it adds more than, let's say, 1% to the compute cost. The reason is we /believe/ there's almost no cheating going on.

Also, it's not the end of the world if a factor is missed. It just means that a P-1 run and then (likely, eventually) two LL tests will be done on the candidate, proving it's composite.

However, I personally feel that /some/ would value having some sort of sanity check on the GPU hardware. As said before, some of us these things for more than just TF'ing (where determinism (or, at least, sanity) is coveted). There would be value in being able to determine if our "kit" was sane while running mfakt[c|o] (as opposed to a GPU based DC test).

Lastly, it would be trivial for the servers to deal with this additional data stream. The heavy-lifting is all in the client (and, thus, the authors of same).
chalsall is offline   Reply With Quote